A reader going by the name of Aint Tellin sent me the following email:

Hello again,

I’ve been watching a number of videos by Dr. Jordan B. Peterson, one of which he stated his IQ. He claimed to have had it tested at one point and that it was “in excess of 150” (although he couldn’t give an exact number). One thing struck me, however. In the same video he mentioned his GRE scores (old GRE) which he listed scoring in the 99th percentile verbally, but only 70th percentile quantitatively.

If I’m not mistaken—although I often am—this would give him a total GRE score somewhere above 1330. While impressive, this only suggests a full-scale IQ around 137. While I have no doubt he’s extremely smart, I believe he only listed his verbal IQ and not his full-scale IQ. I’d like to get your thoughts on this.

Below is a link to the aforementioned video:

 

When asked by a reader what his IQ is, Peterson replies “it’s less than it used to be because it declines as you age”

Actually professional IQ tests like the Wechsler are normed for age, so the average old adult and the average young adult by definition have an IQ of 100.  However Peterson’s correct that the actually number of items correctly answered (raw scores) decline after the 20s, but IQ itself is age controlled.  Maybe Peterson knows all this and is just oversimplifying for the short attention spans on youtube.

I do love what he says about physical exercise staving off cognitive decline though.  In another video he attributes this to the fact that the brain uses tons of oxygen.  Perhaps our resident health expert RR could look into this.

Moving on,  he mentions that he had his IQ tested at one point (why?, when? by who? what test?) and that it was in excess of 150.  My guess is he probably took the WAIS-R while getting his PhD in clinical psychology circa 1990, perhaps as part of the training to administer the test.  If so, the test norms were probably about 12 years old, and assuming James Flynn is right about Wechsler norms inflating at a rate of 0.3 points per year, we may need to deduct 3.6 points (meaning his IQ was in excess of 146 (U.S. norms)).

But that’s speculation on top of speculation.  Let’s turn to his GRE scores.

GRE Verbal  

He mentions that his GRE verbal was in the 99th percentile (which would be at least 2.33 standard deviations (SDs) above the GRE population if we assume their distribution was roughly normal).  Assuming he took the test circa 1984 (when he got his BA), that would have equated to a score of 778+.

We don’t know much how GRE scores equate to IQ, because GREs are normed on aspiring PhDs while IQ is normed on the general U.S. population.  One way to bridge the gap is to convert GRE scores into SAT equivalents, since in rare studies, SATs were taken by the general U.S. population.

A sample of 22,923 people took both the GRE and SAT before 1990.  In this sample, GRE verbal 778 is +2.49 SD (see chart I).

gredata

chart I

To find the verbal SAT equivalent of GRE V 778, we must ask what SAT verbal score is +2.49 with respect to this elite sample.  The answer is 780.

So how does that equate to the IQ scale? We know from a special study, that if all American 17-year-olds had taken the SAT in the 1980s, instead of just the college-bound elite, the average verbal score would have been 376 (SD = 102) (see the The Bell Curve, page 694, note 32). Since by definition, the general U.S. population has an IQ of 100 with an SD of 15, we can infer 376 = IQ 100, and 102 = 15 IQ points.

By this logic, a verbal SAT of 780+ would have equated to an IQ of 159+!

One problem with this method is that it assumes SAT scores are normally distributed at the extremes.

An alternative approach is to look at chart II which equates a GRE V and SAT V of 778+ and 780+ respectively, to IQs 149+ and 156+ respectively.

sat

Chart II (found in the Prometheus MC Reoprt, where it was attributed to Kjeld Hvatum’s  “Letter to Ron Hoeflin” and Ron’s response, In-Genius, # 15, August 1990

Averaging all three estimates gives Peterson a verbal IQ of 155+. This would put Peterson above the one in 8000 level, compared to Americans of his era.

GRE Quant

Peterson claims to have scored in the 70th to 75th percentile on the quant section of the GRE or roughly +0.6 SD above the GRE population if the distribution was normal.  Circa 1984, this equated to a score of 624.

Now if I convert this quant score into an old SAT equivalent, the same way I did for the verbal,  I get 600.

In The Bell Curve they note that if all American 17-year-olds had taken the math SAT in the mid 1980s, the average score (IQ 100) would have been 411 and about the top 0.96% (IQ 135) would have scored 700+.  If we assume a straight line between these data points, a math score of 600 equates to IQ 123.

If so roughly one in 16 Americans of his era are at least as good at the type of math measured by the GRE.

Composite score

Since Peterson’s GRE V and GRE Q equated to old verbal and math SATs of 780+ and 600 respectively, his combined GRE (V + Q) equated to a combined old SAT score of 1380+, which equated to an IQ of about 141+ in Peterson’s day (one in 319 level).

It should be noted however that tests like the GRE and SAT do not market themselves as IQ tests and are designed to predict academic performance, not the general intelligence factor per se.  They also test a narrower and more academic range of brain functions than the Wechsler intelligence scales.